How iOS 15 Live Text Looks Up Word Definitions

Apple’s Live Text feature in iOS 15 utilizes on-device machine learning to recognize text inside images, allowing users to interact with visual words just like native text. By pairing optical character recognition (OCR) with the system-wide "Look Up" feature, iOS 15 enables users to tap and highlight a word in a photo and immediately view dictionary definitions, thesaurus entries, and web-based contextual data.

1. On-Device Text Recognition

When you open an image in the Photos app, Safari, or the Camera viewfinder, the Apple Neural Engine scans the image locally. Deep learning models analyze visual patterns to distinguish letterforms and words from the background. Once identified, these visual elements are converted into selectable, vectorized digital text directly on your device without sending image data to cloud servers.

2. Highlighting and Text Selection

Because the system maps the exact coordinates of the recognized text, the interface allows standard iOS touch gestures. Long-pressing on a word activates selection handles. The operating system treats these boundaries identically to text in a standard document or webpage, presenting the contextual action menu (Copy, Select All, Look Up, Translate, Share).

3. Triggering the Dictionary Query

Tapping Look Up sends the selected string to the native iOS reference system. The operating system checks the word against several internal databases:

• Pre-installed System Dictionaries: Offline reference dictionaries (such as the Oxford Dictionary of English or bilingual dictionaries configured in your device settings) provide grammatical class, pronunciation, and definitions.
• Siri Knowledge: If internet connectivity is available, the system supplements the dictionary entry with structured data from Wikipedia, suggested websites, or media entries related to the word.

4. Displaying the Result Sheet

The results are rendered in a pop-up modal sheet directly over the image. This keeps the user within their current context without forcing them to switch to a web browser or dictionary app. Once closed, the text selection remains active until dismissed, allowing seamless reading and exploration of complex or foreign text captured in photos.